© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 20236
    Prev
    Next
    INDEX
    Explanations

    words related to performance, such as "performance," "penalty," "enhancements," "hardware," "quality," and "throughput."

    oai_token-act-pair · gpt-3.5-turbo

    references to performance metrics and issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 4-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.4.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    4
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    kef
    -0.74
    ean
    -0.69
    een
    -0.69
     Polar
    -0.67
     Narc
    -0.67
    chal
    -0.67
     Griff
    -0.66
    atan
    -0.65
    ertain
    -0.65
    aida
    -0.64
    POSITIVE LOGITS
     optimizations
    1.22
     throughput
    1.10
     optimization
    1.07
     efficiency
    1.00
     acceleration
    0.98
     overhead
    0.95
     debugging
    0.93
     slowdown
    0.93
     bandwidth
    0.91
     consumption
    0.90
    Activations Density 0.129%

    No Known Activations